abstract.computer


Duncan Figurski

The artist and algorithm celebrate the obfuscation of legibility // visibility in joyous and optimistic terms, using abstraction as a playful rebellion against all those things which keep us disconnected from the void.

Instagram
Commercial Design
Collaborative Practice
Videos



CV available upon Request.





BLUE SERIES





What reductionist approach does AI have on our lives? The tools that are used to render data from the world have increased exponentially over the past few years. Machine Vision and Object Detection is now often used to further understand our behavior on a digital scale. There are hard limitations to this software, potentially the most obvious is the fact that abstract or variegated parts of this world are shimied into concise classifications. They are rendered finite, course and marketable. 





This series of paintings began first with the creation of the software.  I wanted to find a system to apply these object classification techniques onto my own abstract pieces, rendering them in a somewhat dystopian reductionism. The software (using Ml5.js, P5.js && Tone.js) would move through each piece in a gridded fashion, isolating one small square at a time and “rendering” it. The classification model I started with was MobileNet.  For each square this software would try and guess what it was looking at: effectively using the classification tools of this oppressive system as a technique to further abstract these pieces. The final result would be the creation of these “product glossaries”. 

This in its own way would become a bizare ritual, I would create a painting based on intuition and that very vulnerable knack would be disected and broken down into the language of a model trained with bias. It would be rendered “understood”, closing the door on the magical liminal space that an abstract piece can take us.  Not abstraction through classification,  but maybe some kind of abstraction from declassification. 







Next I wanted to take the confidence, and pace at which this model was able to classify and make it more visible by using it to play a song. The inner mechanisms of this new tool often go unseen, I wanted to show the gurgling confusion of this “animal” trying to make sense of something I had intentionally designed to be abstract. Like a noisy hog sniffing through a rotting log. 




To engage with this musical interface please click here
            To engage with this musical interface please click here
                        To engage with this musical interface please click here
                                    To engage with this musical interface please click here
                                                To engage with this musical interface please click here
                                                            To engage with this musical interface please click here
                                                                        To engage with this musical interface please click here
                                                                                    To engage with this musical interface please click here
                                                                                                To engage with this musical interface please click here
                                                                                                            To engage with this musical interface please click here
                                                                                                                        To engage with this musical interface please click here
                                                                                                                                    To engage with this musical interface please click here
                                                                                                                                                To engage with this musical interface please click here
                                                                                                                                                          To engage with this musical interface please click here